Archived Forum

You can do what you have proposed but is there anyway of passing a small 2-core wire through the wall?

This way you could just extend the ac adapters lead from the study through the wall and have a emontx v3 located at the consumer unit.

Paul

I did think about that. The problem is I dont think I will be able to find my way though the wall with a cable?

Is that because it's a cavity wall, a tricky route or no hole accessible ?

Are there any existing wires passing through aside from the meter to dist board cables ?

I think a long 10 mm masonry drill is the answer! You won't even need to take the plug off the adapter - always assuming the lead is long enough. At worst you tape the plug onto a length of stiff wire to get it across the cavity. Just make sure you miss the meter tails when drilling.

A drill wont do it...

The problem is the meter is at ground level and the CU is at head height. If I drill through the cable will need to be ducted up the wall in the entrance hallway or the dining room. Both are a no go...

So I really need to 'split' have have 2 emonTx units feeding the emonBase. Is this possbile? How hard is it to set up?

The cheap & simple answer bites the dust then!

It's quite possible for two emonTx's to work into one Base. They won't be synchronised, and it's possible that you might lose the occasional reading if they both happen to try to transmit at the same time.

So you can do as you suggest, but whether the solution will fit your needs and expectations is for you to decide. Here's the problem: If you measure voltage and current separately (for example with two separate multimeters) and then multiply the numbers, you get apparent power. This is exactly the same as the real power measured by your supply meter provided the load is a pure resistance. When you have a reactive load (the motor in a fridge/freezer, for example) the two can be very different.

Your emonTx at your electricity meter can measure the real power, no problem. But the one at the consumer unit can only measure current because it has no access to the voltage measurement. For battery life reasons, it's not possible to have it listening to the voltage that the other one can broadcast, and even if it could, the delay would mean it could not know the phase relationship between current and voltage and so it could only calculate apparent power anyway (and not even the direction of the energy flow!). That would probably not be a major issue, as we can assume the rest of your house loads won't generate although your PV does consume when it's dark, and the power factor of your load (the rest of the grid as seen from your house!) is reasonably well known and reasonably close to unity, so the error won't be that large.

Therefore, your limitation is: you can measure real power (hence energy) imported and exported, but you can only measure current to/from your PV. You'd need to multiply that by the voltage either in your emonGLCD (if you will be having one) or in emonCMS, and do the sums to derive the house consumption and generated power.

But where is your inverter? Can you put a c.t. and an emonTx with its ac-ac adapter near that, and will it be in radio range of your Base? If you can do that, then you know exactly what you PV is doing and with an emonTx by your meter, you know exactly what the nett grid power is doing. You have then what is essentially a Type 2 system but with two emonTx's. The other one to investigate is, does your inverter communicate with the outside world? It might be possible for it to feed data into your system.

Thats a good point about the second Tx not monitoring the voltage. I has assumed that once the voltage was known to the system it was used, not that it was used internally in each Tx locally.

I will need to have a rethink.

I need to have a look inside the consumer unit and see if I can get a clamp on the tails where they come out of the wall into the unit. If I can, I think I may have a spare slot I could connect an external plug to and plug in a single Tx there....

The invter I have has an o/p interface but its a fortune. I've been using a CurrentCost Optismart reader on the generation meter to record my daily output on xively / pvoutput and in my own database to date.

What opportunities are there for monitoring individual appliances?

You can use an optical pulse input into the emonTx of course, but that will probably only give you imported energy (since most meters are programmed to ignore export). If your meter can also output digital data optically, it might be possible to interrogate it: http://openenergymonitor.org/emon/node/626 and this: http://openenergymonitor.org/emon/buildingblocks/interfacing-with-sma-su... is what I had in mind for your inverter. I can't identify yours, so I have no idea what it is capable of.

As far as monitoring individual appliances goes, you can get some information from your overall consumption - check up on "appliance inference". A lot of work is being done on that, I'm not up-to-date on progress. Otherwise, you are talking in terms of an emonTx each, which is going to get expensive. If you know what you're doing with microcontrollers, and with mains electricity from the safety point of view, you might be able to build a single channel monitor cheaper. But if you're not sure, it could end in disaster. I don't know of a cheap commercial offering that we know will interface with the OEM system.

So, a quick update.

Socket added to the consumer unit on a new circuit at the weekend. EmonTx now moved and reinstalled, along with claps for main tail, PV and electric car charger.

Have also added a Temp & Humidity sensor. I have 2 more to add, but have get my head around the whole impenetrable world of arduino....!

Next I have to work out the logic for the feeds config. I did see some pointers on here but have searched and now can't find them!

You create a feed with "log to feed" on the Inputs API page. Only then is your data stored. Prior to "log to feed", you can process the input value - the drop-down list gives the functions available. The raw data is the input to the first function, the output of the first becomes the input to the second, and so on.

Good progress here. All feeds are now 'programmed' and logging. I worked out the maths OK!

Next is the arduino stuff so I can set unique numbers for each of my temperature nodes. I started to have a look last night but am trying to get my head round all the code etc and github.

You don't really need to worry about Github. You can instead download zip files and use those. (IMHO, if you're not likely to want to update regularly, this is an easier and better way.)

All you need is the sketch for your emonTx, change one line: const int nodeID = 10; and reload the sketch.